Mino1289
Mino1289
I used the script and quantization method here : https://gist.github.com/sayakpaul/b664605caf0aa3bf8585ab109dd5ac9c The script by AmericanPresidentJimmyCarter.
As I read the issue and PR you linked, the issue i'm facing is most likely due to quanto not being supported with peft. Using BitsAndBytesConfig should bypass the problem,...
And in 8 bit ? My issue is about 8 bit of qfloat8. It's okay for the name. Thanks for the quick support!
And diffusers 0.32.1 or from source ?